Exploring structural and electronic effects in three isomers of tris{bis(trifluoromethyl)phenyl}borane: towards the combined electrochemical-frustrated Lewis pair activation of H₂

Abstract

Three structural isomers of tris{bis(trifluoromethyl)phenyl}borane have been studied as the acidic component of frustrated Lewis pairs. While the 3,5-substituted isomer is already known to heterolytically cleave H₂ to generate a bridging-hydride; ortho-substituents in the 2,4- and 2,5-isomers quench such reactivity through electron donation into the vacant boron p_z orbital and steric blocking of the boron centre; as shown by electrochemical, structural and computational studies. Electrochemical studies of the corresponding borohydrides identify that the two-electron oxidation of terminal-hydrides occurs at more positive potentials than observed for [HB(C₆F₅)₃]⁻, while the bridging-hydride oxidizes at a higher potential still, comparable to that of free H₂.

---

Introduction

Since the pioneering work of Stephan's group¹ the field of frustrated Lewis pair (FLP) chemistry has grown rapidly.^2–7 The archetypal FLP system combines the Lewis acid B(C₆F₅)₃ with a sterically demanding Lewis base such as P(^tBu)₃, to effect the heterolytic cleavage of H₂, resulting in the formation of the corresponding hydridic and protic products respectively.

FLPs have found applications as catalysts or mediators for a variety of reactions such as the metal-free hydrogenation of imines and nitriles,⁸ alkynes,⁹ silyl enol ethers,¹⁰ and ketones.^11,12 The activation of small molecules such as CO₂,^13,14 and alkynes by FLPs has also found applications in the synthesis of heterocycles and other aromatic systems.¹⁵

While much of the literature has focused on tris(pentafluorophenyl)borane, B(C₆F₅)₃ as the Lewis acidic component in FLPs, other electron-deficient boranes have been used, including a range of halogenated triarylboranes,^14,16–18 and borenium cations,^19,20 Other examples of Lewis acids that have found use in FLPs include: the triaryl aluminium species, Al(C₆F₅)₃, which generates a bridging-hydride following cleavage of H₂;²¹ and the carbon based N-methylacridinium salts which activate H₂ even in the presence of H₂O.²² FLPs are not limited to the main group, with zirconocene–phosphane complexes pioneered by Wass and co-workers shown to act as intramolecular FLPs that exhibit unprecedented reactivity towards small molecules;²³ similar chemistry has also been demonstrated with zirconocene-amines which act as hydrogenation catalysts with a wide range of substrates.²⁴

In 2014 we introduced the concept of “combined electrochemical-frustrated Lewis pairs”,^20,25,26 that couple the heterolytic cleavage of H₂ by a conventional FLP with in situ electrochemical oxidation of the resultant borohydride and subsequent regeneration of the parent borane. The combined electrochemical-FLP systems were shown to be electrocatalytic for the oxidation of H₂ (to form two protons and two electrons – a key reaction in many hydrogen-based energy technologies). These preliminary reports represent the first application of FLP chemistry other than to catalyse the hydrogenation of small molecules. For example, by combining the archetypal B(C₆F₅)₃/P(^tBu)₃ FLP in the presence of H₂ whilst at the same time oxidizing the resultant [HB(C₆F₅)₃]⁻ intermediate formed, the potential for oxidation (the energetic driving force) of the borohydride at a glassy carbon electrode was found to be reduced by 0.61 V, in comparison to the direct oxidation of H₂ (observed at ca. +1.49 V vs. [FeCp₂]^0/+ under the same conditions), an energy saving equivalent to 117.7 kJ mol⁻¹.²⁵

Herein we report studies on three isomers of tris{bis(trifluoromethyl)phenyl}borane (Fig. 1) and their associated tris{bis(trifluoromethyl)phenyl}borohydrides. Electrochemical and computational investigations explain their varying ability to heterolytically cleave H₂ as part of a FLP and allow us to extend our studies into combined electrochemical-FLPs.

Fig. 1 Isomers of tris{bis(trifluoromethyl)phenyl}borane.

Results and discussion

Synthesis and characterization

The synthesis of tris{3,5-bis(trifluoromethyl)phenyl}borane 1 has been previously reported,^18,27 and whilst a synthetic route to tris{2,4-bis(trifluoromethyl)phenyl}borane 2 has been reported by Cornet et al.²⁸ their method led to a mixture of products requiring the isolation of 2 in low yield by fractional sublimations. The commercial availability of 2,4-bis(trifluoromethyl)bromobenzene and 2,5-bis(trifluoromethyl)bromobenzene allows us to report an improved synthetic route to 2 and also the synthesis of the novel tris{2,5-bis(trifluoromethyl)phenyl}borane 3.

Boranes 2 and 3 were synthesized by lithium–halogen exchange of BrC₆H₃(CF₃)₂ with ⁿBuLi at −77 °C generating the aryllithium species LiC₆H₃(CF₃)₂. Treatment of this with a third of an equivalent of BCl₃, followed by warming to room temperature, subsequent removal of the volatiles and purification by sublimation or recrystallization allowed for the isolation of pure 2 and 3, in 90 and 77% yields respectively. It should be noted that applying the same synthetic route for the synthesis of 1, leads to a mixture of products with [Li(OEt₂)_n][B{3,5-(CF₃)₂C₆H₃}₄] as the major component.

The crystal structures of 1²⁹ and 2²⁸ have been previously reported. Single crystals of 3 were obtained by slow diffusion of a saturated CH₂Cl₂ solution of 3 into n-hexane at −25 °C, from which the crystal structure was obtained in collaboration with the EPSRC UK National Crystallography Service³⁰ (Fig. 2 and S1,† and Table 2).

Fig. 2 The crystallographic molecular structure of B{2,5-(CF₃)₂C₆H₃}₃3 (hydrogen atoms removed for clarity).

The structures of 1–3 all show similar features, with a trigonal-planar boron centre and the three aryl rings twisted with respect to the BC₃ plane to minimize steric interactions between the aryl rings. The degree of twist can be quantified by an appropriate choice of C–B–C–C torsion angles, which are both smaller and more consistent for 1 {mean 36(2)°, range 33.2–38.9°} than for 2 {mean 54(11)°, range 42.3–68.3°} or 3 {mean 53(9)°, range 40.9–61.2°}, due to the steric bulk of ortho-trifluoromethyl substituents. For 2 and 3 this results in the ortho-CF₃ groups orientated above/below the boron centre resulting in B⋯F distances of 2.81(1), 2.80(7) Å for 2 and 3 respectively. The orientation of these groups suggests the possibility of electron donation from the fluorine atoms into the formally vacant p_z orbital at boron; an effect together with the steric influence of the ortho-CF₃ groups which would be expected to have a significant influence on the Lewis acidity/reactivity of these compounds. An example of the type of reactivity which could be expected from the ortho-CF₃ groups, was observed by Cornet et al., who reported evidence of B–Cl/B–F exchange in mixtures containing the boranes B(Ar^F)₂Cl where Ar^F = 2,4,6-(CF₃)₃C₆H₂, 2,4-(CF₃)₂C₆H₃ or 2,6-(CF₃)₂C₆H₃.²⁸

The reactivity of 1 as the Lewis acidic component of an FLP has been previously studied,¹⁸ which showed that the reaction of 1 with H₂ in the presence of 2,2,6,6-tetramethylpiperidine (tmp) leads to rapid formation of the bridging hydride species [tmpH][(μ-H)(1)₂]. Analogous reactions of 2 and 3 with H₂ in the presence of the Lewis bases tmp or P(^tBu)₃, result in neither Lewis acid–base adduct formation nor any evidence of H₂ activation observable by NMR spectroscopy over a minimum period of 48 hours.

Direct synthesis of authentic terminal-hydride species [HB{C₆H₃(CF₃)₂}₃]⁻ proved to be possible for all three isomers 1–3 by direct reaction of the borane with sodium triethylborohydride in toluene solution, resulting in near quantitative conversion to the corresponding borohydrides. NMR spectra of the terminal-hydride species Na[1-H]–Na[3-H] show doublets in the ¹¹B spectra (δ_B: −9.1, −15.3, −14.2 and ¹J_BH: 88, 93, 84 Hz respectively) and broad 1 : 1 : 1 : 1 quartets in the proton spectra (δ_H: +3.66, +4.06, +3.06 respectively). Additionally, for Na[2-H] and Na[3-H] clear evidence of through-space coupling between the hydride and ortho-CF₃ groups is observed in the ¹⁹F spectra (J_FH = 6.8, 7.6 Hz respectively); further, in the case of Na[3-H] this coupling is also observable in the proton spectrum with the hydride signal observable as a 1 : 1 : 1 : 1 quartet of broad 1 : 3 : 3 : 1 quartets (Fig. 3).

Fig. 3 Hydride resonance in ¹H NMR spectrum of Na[3-H], showing coupling to both ¹¹B (I = 3/2) and an ortho-CF₃ group {3× ¹⁹F, I = 1/2).

Synthesis of Na[2-H] and Na[3-H] allows for the confirmation that the inability of the 2,3/P(^tBu)₃ FLPs to cleave H₂ is not due to unfavourable thermodynamics, but due to the significant kinetic barrier resulting from the steric and electronic effects of the ortho-CF₃ groups. Reaction with authentic [(^tBu)₃PH]Cl results in rapid metathesis (indicated by precipitation of NaCl) and formation of the salts [(^tBu)₃PH][2-H] and [(^tBu)₃PH][3-H]. NMR spectra of which, show no liberation of H₂, regeneration of free borane/phosphine, or any other evidence of reaction over a 66 hour period.

Electrochemical studies

Cyclic voltammetric studies of 1–3 were performed in the non-coordinating solvent CH₂Cl₂ using [ⁿBu₄N][B(C₆F₅)₄] as the added electrolyte at a glassy carbon electrode (GCE). In all three cases a one-electron, reduction process is observed with quasi-reversible (moderately fast) electron transfer kinetics. However for 1 (Fig. 4a) the process appears to be chemically irreversible (but not electrochemically irreversible) at scan rates up to 2.0 V s⁻¹ due to fast chemical follow-up kinetics cf. the rate of electron transfer, while for both 2 (Fig. 4b) and 3 (Fig. 4c) an associated oxidation wave is observed indicating that the radical anions 2˙⁻ and 3˙⁻ are sufficiently stable on the electrochemical timescale for their subsequent (re-)oxidation to be observable.

Fig. 4 Experimental (line) and simulated (open circles) cyclic voltammograms for the reduction of (a) 1, (b) 2, and (c) 3.

Based on our previous experience from an electrochemical study of B(C₆F₅)₃,³¹ we performed digital simulations of the experimental voltammetric data modelled using an EC-mechanism (i.e. a reversible, heterogeneous electron transfer step followed by an irreversible, homogeneous chemical step which generates electro-inactive products. Other postulated mechanisms produced a poor fit to the data). These digital simulations allowed us to extract pertinent mechanistic parameters such as the formal redox potentials and charge transfer coefficients (E° and α respectively) and kinetic parameters for the electron transfer (k°) and follow-on chemical step (k_f) as shown in Table 1.

Table 1 Mechanistic parameters obtained by digital simulation of voltammetric data for the one-electron reductions of 1–3

		1	2	3
a Diffusion constants (D) obtained via^1H and ^19F DOSY NMR spectroscopy. b k f values are modelled as a pseudo first-order process.
BArF18 + e^− = BArF18˙^−	E° vs. [FeCp2]^0/+/V	−1.61 ± 0.01	−1.79 ± 0.01	−1.85 ± 0.01
	α	0.419	0.498	0.468
	k°/cm s^−1	4.56 × 10^−3	2.61 × 10^−2	2.22 × 10^−2
BArF18˙^− ⇒ ‘decomposition’	k f/s^−1 ^b	≥25	1.92 × 10^−2	8.08 × 10^−2
D(BArF18) = D(BArF18˙^−)/cm^2 s^−1 ^a	3.76 × 10^−5	1.13 × 10^−5	1.13 × 10^−5

---

Table 2 Crystallographic data for 3

	B{2,5-(CF3)2C6H3}33
Empirical formula	C24H9BF18
Formula weight	650.12
Temperature/K	100
Crystal system	Triclinic
Space group	P1̄
a/Å	7.2951(5)
b/Å	10.6358(7)
c/Å	15.9794(11)
α/°	85.950(4)
β/°	86.582(4)
γ/°	74.846(4)
Volume/Å^3	1192.64(14)
Z	2
ρ calc/mg mm^−3	1.810
μ/mm^−1	0.204
F(000)	640.0
Crystal size/mm^3	0.30 × 0.06 × 0.02
Radiation	Mo Kα (λ = 0.71075 Å)
2Θ range for data collection	3.974 to 54.96°
Index ranges	−9 ≤ h ≤ 8, −13 ≤ k ≤ 13, −20 ≤ l ≤ 20
Reflections collected	17 457
Independent reflections	5459 [Rint = 0.0820, Rsigma = 0.0575]
Data/restraints/parameters	5459/0/388
Goodness-of-fit on F^2	1.055
Final R indexes [I ≥ 2σ(I)]	R 1 = 0.0555, wR2 = 0.1508
Final R indexes [all data]	R 1 = 0.0712, wR2 = 0.1631
Largest diff. peak/hole/e A^−3	0.56/−0.39

---

The formal reduction potentials (E°) suggest that 1 is the most electrophilic of the three boranes, while 3 is the least. While the difference in electrophilicity between 2 and 3 is consistent with a simplified view of the inductive electron withdrawing effects of the different meta- and para-(CF₃) group positions based on Hammett parameters {σ_meta(CF₃) = 0.43 vs. σ_para(CF₃) = 0.54}³² the relative electrophilicity of 1 cannot be similarly rationalized. This is in part because simple Hammett parameters for ortho-(CF₃) substituents do not satisfactorily account for any additional steric and/or electronic effects. We can rationalize the reduced electrophilicity of 2 and 3 in comparison to 1 by considering the electronic effect of having ortho-CF₃ groups present in 2 and 3. As noted above, the crystal structures of 2 and 3 show the ortho-(CF₃) groups are positioned at sufficiently close distances above the central BC₃ plane such that donation from the lone pairs on the fluorine atoms into the vacant boron p_z orbital on boron could occur; such donation of electron density would be expected to reduce the electrophilicity of the boron centre, as is observed experimentally by our electrochemical measurements and indicated in electronic structure DFT calculations (vide infra).

Furthermore, the steric shielding of the boron centre by the ortho-(CF₃) groups in 2 and 3 is qualitatively evident from the observation of more reversible redox processes in the cyclic voltammetric data. Quantitatively this is shown in the values of the rate constants (k_f) obtained from voltammetric digital simulation for the radical anion decomposition step, which is assumed to proceed in a similar fashion as was previously ascertained for the analogous [B(C₆F₅)₃]˙⁻ intermediate via reaction between solvent molecules and the boron centre in the reduced radical anion intermediates, 1˙⁻, 2˙⁻, or 3˙⁻.³¹ The value of k_f is at least three orders of magnitude greater for the decomposition of 1˙⁻ than for 2˙⁻ or 3˙⁻, where, in the latter two cases, the presence of ortho-CF₃ groups provides significant steric shielding to the boron centre. It is worth noting here that these findings demonstrate the ease with which synthetic chemists working in this area can gain powerful insights into the chemistry of Lewis acidic species by the application of simple, rapid electrochemical characterization techniques in addition to the more ubiquitous crystallographic and spectroscopic characterization techniques. Simple examination of the shape and position of the voltammetry of each borane, obtained in a 20 minute experiment using inexpensive equipment can tell us qualitatively that in comparison to 1, boranes 2 and 3 are more sterically hindered, and less electrophilic, and therefore less Lewis acidic and less likely to be active FLP components for H₂ activation. The voltammetry even allows us to infer why this is so, given that the only thing boranes 2 and 3 have that 1 does not have and that could simultaneously sterically shield the borane and reduce the electronic demand at the boron centre are the o-CF₃ groups.

In comparison with the archetypal Lewis acid, B(C₆F₅)₃ {which is observed under the same conditions as a quasi-reversible reduction at E° = −1.518 V vs. [FeCp₂]^0/+ (see Fig. S2†)}, boranes 1–3 are all less electrophilic; while [B(C₆F₅)₃]˙⁻ is more stable (smaller k_f) with respect to follow-on decomposition reactions than 1˙⁻, but still considerably less stable than the ortho-(CF₃) stabilized 2˙⁻ and 3˙⁻.

Borohydride oxidation. Our combined electrochemical-FLP concept requires the electrochemical oxidation of borohydrides generated by the heterolytic cleavage of H₂. Despite the only such species generated from the B{C₆H₃(CF₃)₂}₃ isomers, being the previously reported bridging hydride [(μ-H)(1)₂]⁻,¹⁸ the ease of direct synthesis of the terminal hydrides [HB{C₆H₃(CF₃)₂}₃]⁻ allows for their electrochemical study, to add to our understanding of the electrochemistry of the triarylborohydrides. Therefore, electrochemical studies of Na[1-H]–Na[3-H] and [tmpH][(μ-H)(1)₂] were performed under the same conditions as their borane precursors (although Na[1-H] and [tmpH][(μ-H)(1)₂] proved only sparingly soluble in CH₂Cl₂, and hence their concentrations cannot be accurately determined). In all cases a single irreversible oxidation wave is observed at scan rates up to 2.0 V s⁻¹.

For the terminal hydrides Na[1-H]–Na[3-H] (Fig. 5a–c) oxidations occurred with peak potentials of +1.08, +1.31, +1.13 V vs. [FeCp₂]^0/+ at a scan rate of 100 mV s⁻¹ respectively; with no evidence of electroactive product species (such as the parent boranes) being regenerated in sufficient quantities to be observed.

Fig. 5 Experimental cyclic voltammograms for the oxidation of (a) Na[1-H], (b) Na[2-H], and (c) Na[3-H] shoulder (*) visible at the higher scan rates due to trace impurity in the solvent/electrolyte.

The bridging hydride [tmpH][(μ-H)(1)₂] (Fig. 6) oxidation is observed as a shoulder on the edge of the solvent window, at ca. +1.55 V vs. [FeCp₂]^0/+ at a scan rate of 100 mV s⁻¹ (observable distinct to the solvent/electrolyte breakdown at scan rates below 1.0 V s⁻¹), trace amounts of the terminal hydride species [1-H]⁻ are also observed. Following oxidation of [(μ-H)(1)₂]⁻, sweeping to negative potentials results in observation (at 100 mV s⁻¹ scan rate and above) of an irreversible reduction wave at ca. −1.7 V vs. [FeCp₂]^0/+ characteristic of the reduction wave observed for the parent borane 1.

Fig. 6 Experimental cyclic voltammogram(s) for the oxidation of [tmpH][(μ-H)(1)₂].

In comparison to our previous studies on [HB(C₆F₅)₃]⁻,²⁵ all of the terminal-hydrides oxidize at more positive potentials, whilst the bridging-hydride species, [(μ-H)(1)₂]⁻, is oxidized at an even more positive potential approaching that of the direct oxidation of H₂ at a GCE (ca. +1.5 V vs. [FeCp₂]^0/+), and is effectively behaving as an electrolyte under these conditions.

Computational studies

To further our understanding of the tris{bis(trifluoromethyl)phenyl}borane isomers 1–3 we have investigated them and their associated radical-anions 1˙⁻–3˙⁻ using density functional theory (DFT).

For all the neutral boranes 1–3 the LUMOs (Fig. 7) showed a high degree of boron p_z character (19, 22, 19% respectively, the majority of the remaining contributions being from the phenyl π-systems) as do the SOMOs (Fig. S4†) for the radical-anions 1˙⁻–3˙⁻.

Fig. 7 Lowest unoccupied molecular orbitals (LUMOs) for (a) 1, (b) 2, and (c) 3.

Total molecular energy calculations indicate that the ground state of 1 is lower in energy than those of 2 and 3 by ca. 43 kJ mol⁻¹. Analysis of calculated Mulliken atomic charges for 1–3, show that any stabilisation of 1 compared to 2 and 3 cannot be attributed to electron withdrawing effects of the aryl rings; and therefore may be attributed to the reduced steric hindrance caused by the lack of ortho-CF₃ groups in 1.

As noted previously, the ortho-CF₃ groups in 2 and 3 are orientated such that there is potential for B⋯F bonding interactions. Such interactions are clearly identified by the calculated bonding parameters: the B⋯F bonding parameter for 2 is 0.065, whilst for 3 it is 0.076 (in both cases averaged over all contributing B⋯F pairs). As expected, in 1, where no interaction occurs, the B⋯F bonding parameter to the meta-CF₃ groups is <0.001.

Conclusions

We have investigated three structural isomers of tris{bis(trifluoromethyl)phenyl}borane in terms of both their electrochemical redox chemistries and also as the Lewis acidic components of an FLP for the heterolytic cleavage of H₂.

The two isomers of tris{bis(trifluoromethyl)phenyl}borane that incorporate ortho-CF₃ groups were not found to be active as the Lewis acidic component of FLPs for H₂ cleavage reactions. This lack of reactivity is due to a combination of, kinetic effects resulting from in part steric shielding of the boron centre; but also quenching of the boranes’ electrophilicity through B⋯F bonding interactions, which are quantified by DFT calculations and electrochemical measurements. Whilst electrochemical studies show that all three isomers are less electrophilic than the archetypal Lewis acid B(C₆F₅)₃ the link between Lewis acidity, activity towards H₂ in an FLP, and electrophilicity, as shown in our previous studies^20,25,26 and those of other groups^33–35 is complex and requires further study.

The direct synthesis of all three tris{bis(trifluoromethyl)phenyl}borohydride species under mild conditions allows for their reduction potentials to be measured, and their propensity for combined electrochemical-frustrated Lewis pair catalysis to be screened in a straightforward manner. The oxidation potentials of all three terminal borohydrides studied were found to be more positive than that of [HB(C₆F₅)₃]⁻ yet are still less than the potential required for the direct oxidation of H₂ at a GCE under identical conditions.

Electrochemical studies of the bridging hydride formed when H₂ is cleaved by the B{3,5-(CF₃)₂C₆H₃}₃/tmp FLP, show that this species is oxidized at comparable potentials to that of the direct oxidation of H₂. However, following oxidation, the regeneration of the parent borane species is clearly observed, which is not the case for any of the terminal borohydride Na[1-H]–Na[3-H] species studied.

What this report demonstrates is the utility of electrochemical characterization methods to enable synthetic chemists to rapidly screen prospective new Lewis acids using simple electrochemical techniques to gain insights into the chemical behaviour of new species. It also provides insights to guide the design of new Lewis acids for researchers wishing to employ the combined electrochemical-frustrated Lewis pair activation of H₂, which is the focus of our ongoing studies.

Experimental

All reactions and manipulations were performed under an atmosphere of dry, oxygen-free N₂, using either standard Schlenk techniques or in either a MBraun UNIlab or LABmaster glovebox. All solvents were dried prior to use by refluxing over an appropriate drying agent {Na/benzophenone for petroleum ether (b.p 40–60 °C) and diethyl ether; Na for toluene; CaH₂ for dichloromethane}, collected by distillation under an inert N₂ atmosphere and stored over 4 Å molecular sieves prior to use. All other reagents were obtained from commercial suppliers and used as received.

NMR Spectra were obtained on either a Bruker Avance III 500 MHz or Bruker AV 400 MHz spectrometer, all deuterated solvents were dried over 4 Å molecular sieves prior to use. For ¹H spectra residual protio-solvent was used as an internal standard; for ¹³C the solvent resonance(s) were used as an internal standard;³⁶ for ¹⁹F spectra CFCl₃ was used as an external standard; for ¹¹B spectra BF₃·Et₂O was used as an external standard. ¹H and ¹⁹F DOSY experiments were performed on a Bruker Avance III 500 MHz spectrometer equipped with a broadband multinuclear probe, using a longitudinal eddy current delay incorporating bipolar gradients for diffusion and spoil gradients (ledbpgp2s) pulse sequence.³⁷

Mass spectrometry was performed by the EPSRC Mass Spectrometry Service at the University of Swansea for 3, or by Dr L. Haigh at Imperial College using a Micromass Autospec Premier spectrometer for 2. Elemental analyses were performed by Mr S. Boyer of the Elemental Analysis Service at London Metropolitan University.

Single crystals of 3 were grown by slow diffusion of a saturated CH₂Cl₂ solution of the compound into n-hexane; data collection and processing was performed at the UK National Crystallographic Service at the University of Southampton.³⁰ Using Olex2,³⁸ the structure was solved and space group assigned with SuperFlip/EDMA³⁹ using charge flipping, and then refined with the ShelXL version 2014/7⁴⁰ refinement program using least squares minimization.

CCDC 1061234 contains the supplementary crystallographic data for this paper.

Electrochemical studies were carried out using a Metrohm Autolab PGSTAT302N potentiostat linked to a computer running Metrohm Autolab NOVA version 1.11 software, in conjunction with a three electrode cell comprising: a glassy carbon disc working electrode (Bioanalytical Systems Inc., ca. 7.0 mm² area calibrated using the [FeCp₂]^0/+ redox couple), a platinum wire (99.99% purity) counter electrode, and a silver wire (99.99% purity) pseudo-reference electrode; all electrodes were polished with 0.3 μm α-alumina and dried prior to use. All electrochemical measurements were performed at ambient temperature under a dry N₂ atmosphere, in CH₂Cl₂ containing 0.05 M [ⁿBu₄N][B(C₆F₅)₄] as the supporting electrolyte and between 1.0 and 2.0 mM of the analyte species of interest. Cyclic voltammetric measurements were iR-compensated using positive-feedback to within 85 ± 5% of the uncompensated solution resistance. [ⁿBu₄N][B(C₆F₅)₄] was synthesized according to published methods.⁴¹ All potentials were referenced to the [FeCp₂]^0/+ redox couple, which was added as an internal standard. Simulations of electrochemical processes were performed using ElchSoft DigiElch version 7.096 software.⁴²

DFT calculations were performed using the Gaussian 09 computational package.⁴³ Geometry optimization calculations have been carried out using the three-parameter exchange functional of Becke⁴⁴ (B3) and the correlation functional of Lee, Yang, and Parr (LYP), B3LYP.⁴⁵ The 6-311+G(d,p) basis set has been implemented for all atoms.⁴⁶ Structures were geometry optimized in the gas phase with the default convergence criteria and confirmed as minima through frequency calculations. All optimized structures were confirmed as minima by frequency analysis with thermodynamic properties extracted for the gas phase at 298.15 K and 1 atm. Bonding parameters between B and F atoms were calculated as the absolute values of the associated non-diagonal elements of the condensed to atoms electron density matrix. Density matrixes were computed in a separate calculation taking into consideration basis set superposition error (BSSE) correction using an unrestricted Hartree–Fock (H–F) calculation with counterpoise (CP) correction approach as implemented in Gaussian suite.

B{3,5-(CF₃)₂C₆H₃}₃1^18,27 and [tmpH][(μ-H)(B{3,5-(CF₃)₂C₆H₃}₃)₂] [tmpH][(μ-H)(1)₂]¹⁸ were synthesized as previously reported.

B{2,4-(CF₃)₂C₆H₃}₃2

2,4-Bis(trifluoromethyl)bromobenzene (2.00 g, 1.16 cm³, 6.83 mmol) and Et₂O (100 cm³) were combined and the solution cooled to −77 °C. With the aid of rapid stirring, ⁿBuLi (2.87 cm³, 7.17 mmol, 2.5 M in hexanes) was added slowly by means of a syringe. Following one hour of stirring, BCl₃ (2.28 cm³, 2.28 mmol, 1.0 M in heptane) was syringed into the amber solution and the mixture permitted to warm slowly to room temperature. The volatiles were removed in vacuo and the off white residue extracted with CH₂Cl₂ (3 × 25 cm³) and filtered through Celite. Volatiles were removed under vacuum, and following a high vacuum sublimation step (10⁻⁶ mbar) at 85 °C, a pure white solid was obtained. Yield 1.33 g (2.04 mmol, 90%).

¹H NMR (400.4 MHz, CD₂Cl₂, 25 °C, δ): +8.06 (s, 3H, 3-H), +7.87 (d, 3H, ³J_HH = 8 Hz, 5-H), +7.46 (d, 3H, ³J_HH = 8 Hz, 6-H); ¹¹B NMR (128.4 MHz, CD₂Cl₂, 25 °C, δ): +74.0 (br.s); ¹³C{¹H} NMR (100.6 MHz, CD₂Cl₂, 25 °C, δ): +144.2 (br, 1-C), +135.9 (s, 6-CH), +134.2 (q, ²J_CF = 34 Hz, 2/4-C), +133.7 (q, ²J_CF = 34 Hz, 2/4-C), +127.9 (q, ³J_CF = 3 Hz, 5-CH), +123.9 (q, ¹J_CF = 273 Hz, 2/4-CF₃), +123.6 (sept., ³J_CF = 3 Hz, 3-CH), +123.6 (q, ¹J_CF = 273 Hz, 2/4-CF₃). ¹⁹F NMR (376.8 MHz, CD₂Cl₂, 25 °C, δ): −56.6 (s, 9F, 2-CF₃), −63.8 s, 9F, 4-CF₃). HRMS-EI (m/z): [M]⁺ calc. for C₂₄H₉BF₁₈, 650.0510; found, 650.0491. Elemental analysis (calc. for C₂₄H₉B₁F₁₈): C 44.34 (44.48), H 1.40 (1.47).

B{2,5-(CF₃)₂C₆H₃}₃3

2,5-Bis(trifluoromethyl)bromobenzene (3.1 cm³, 17.9 mmol) and Et₂O (100 cm³) were combined and cooled to −77 °C. ⁿBuLi (11 cm³, 17.6 mmol, 1.6 M in hexanes) was added to the stirred solution. After one hour BCl₃ (5.8 cm³, 5.8 mmol, 1.0 M in heptane) was added to the orange solution and the mixture permitted to slowly warm to room temperature. The volatiles were removed in vacuo and the pale yellow residue extracted with CH₂Cl₂ (3 × 25 cm³) and filtered (via cannula). The product is then recrystallized from CH₂Cl₂/n-hexane and isolated as a white micro-crystalline solid. Yield 2.89 g (4.44 mmol, 77%).

¹H NMR (500.21 MHz, CD₂Cl₂, 25 °C, δ): +7.95 (s, 3H, 3/4-H), +7.95 (s, 3H, 3/4-H), +7.47 (s, 3H, 6-H); ¹¹B NMR (160.49 MHz, CD₂Cl₂, 25 °C, δ): +70.7 (br.s); ¹³C{¹H} NMR (125.78 MHz, CD₂Cl₂, 25 °C, δ): +141.5 (br.s, 1-C), +137.1 (q, ²J_CF = 33 Hz, 2/5-C), +133.3 (q, ²J_CF = 33 Hz, 2/5-C), +132.4 (q, ³J_CF = 3.7 Hz, 3/4/6-C), +129.3 (q, ³J_CF = 3.7 Hz, 3/4/6-C), +127.8 (br.m, 3/4/6-C), +124.1 (q, ¹J_CF = 275 Hz, 2/5-CF₃), +123.9 (q, ¹J_CF 273 Hz, 2/5-CF₃); ¹⁹F NMR (470.67 MHz, CD₂Cl₂, 25 °C, δ): −56.5 (s, 9F, 2-CF₃), −63.7 (s, 9F, 5-CF₃). HRMS-APCI (m/z): [M − F]⁺ calc. for C₂₄H₉BF₁₇, 631.0525; found, 631.0519. Elemental analysis (calc. for C₂₄H₉B₁F₁₈): C 44.22 (44.48), H 1.38 (1.47).

Na[HB{3,5-(CF₃)₂C₆H₃}₃] Na[1-H]

To a solution of 1 (0.40 g, 0.62 mmol) in toluene (10 cm³) was added Na[HBEt₃] (0.6 cm³, 0.6 mmol, 1.0 M in toluene), the reaction mixture was stirred for 6 hours to give a colourless solution. All volatiles were removed in vacuo, to give a white residue which was washed with petroleum ether (2 × 5 cm³) and dried in vacuo to give a white solid. Yield 0.40 g (0.59 mmol, 95%).

¹H NMR (500.21 MHz, CD₃CN, 25 °C, δ): +7.70 (s, 6H, 2,6-H), +7.58 (s, 3H, 4-H), +3.66 (br.q, ¹J_HB = 84 Hz, 1H, HB); ¹¹B NMR (160.49 MHz, CD₃CN, 25 °C, δ): −9.1 (d, ¹J_BH = 88 Hz); ¹³C{¹H} NMR (125.78 MHz, CD₃CN, 25 °C, δ): +165.0 (q, ¹J_CF = 49 Hz, 1-C), +135.7 (s, 2,6-C), +130.1 (q, ²J_CF = 32 Hz, 3,5-C), +126.0 (q, ¹J_CF = 272 Hz, CF₃), +118.5 (s, 4-C); ¹⁹F NMR (470.67 MHz, CD₃CN, 25 °C, δ): −63.0 (s, 18F, CF₃). Elemental analysis (calc. for C₂₄H₁₀B₁F₁₈Na): C 42.95 (42.76), H 1.61 (1.50).

Na[HB{2,4-(CF₃)₂C₆H₃}₃] Na[2-H]

To a solution of 2 (0.275 g, 0.42 mmol) in toluene (8 cm³) was added Na[HBEt₃] (0.43 cm³, 0.43 mmol, 1.0 M in toluene), the reaction mixture was stirred for 4 hours to give a cloudy white suspension. The reaction mixture was concentrated in vacuo reducing its volume to ca. 2 cm³, and the product precipitated by addition of petroleum ether and cooling to −25 °C. The white solid was isolated by filtration and dried in vacuo. Yield 0.231 g (0.34 mmol, 81%).

¹H NMR (500.21 MHz, CD₃CN, 25 °C, δ): +7.76 (s, 3H, 3-H), +7.47 (d, ³J_HH = 7.9 Hz, 3H, 5-H), +7.11 (d, ³J_HH = 7.9 Hz, 3H, 6-H), +4.06 (br.q, ¹J_HB = 93 Hz, 1H, BH); ¹¹B NMR (160.49 MHz, CD₃CN, 25 °C, δ): −15.33 (d, ¹J_BH = 93 Hz); ¹³C{¹H} NMR (125.78 MHz, CD₃CN, 25 °C, δ): +139.1 (s, 6-CH), +135.3 (q, ²J_CF = 29 Hz, 2/4-C), +126.7 (q, ³J_CF = 3.7 Hz, 5-CH), +126.7 (q, ¹J_CF = 275 Hz, 2/4-CF₃), +126.3 (q, ²J_CF = 32 Hz, 2/4-C), +126.2 (q, ¹J_CF = 272 Hz, 2/4-CF₃), +122.6 (sept., ³J_CF = 3.7 Hz, 3-CH),; ¹⁹F NMR (470.67 MHz, CD₃CN, 25 °C, δ): −59.5 (d, J_FH = 6.8 Hz, 9F, 2-CF₃), −62.6 (s, 9F, 4-CF₃). Elemental analysis (calc. for C₂₄H₁₀B₁F₁₈Na): C 42.83 (42.76), H 1.57 (1.50).

Na[HB{2,5-(CF₃)₂C₆H₃ }₃] Na[3-H]

To a solution of 3 (0.275 g, 0.42 mmol) in toluene (8 cm³) was added Na[HBEt₃] (0.43 cm³, 0.43 mmol, 1.0 M in toluene), the reaction mixture was stirred for 4 hours to give a colourless solution. The reaction mixture was concentrated in vacuo reducing its volume to ca. 2 cm³, and the product crystallized by addition of petroleum ether and cooling to −25 °C. The white crystalline solid was isolated by filtration and dried in vacuo. Yield 0.207 g (0.31 mmol, 74%).

¹H NMR (500.21 MHz, CD₂Cl₂, 25 °C, δ): +7.70 (d, ³J_HH = 8.2 Hz, 3H, 3/4-H), +7.45 (d, ³J_HH = 8.5 Hz, 3H, 3/4-H), +7.23 (br.s, 3H, 6-H), +3.06 (br.qq, ¹J_HB = 82 Hz, J_HF ≈ 9.1 Hz, 1H, BH); ¹¹B NMR (160.49 MHz, CD₂Cl₂, 25 °C, δ): −14.23 (d, ¹J_BH = 84 Hz); ¹³C{¹H} NMR (125.78 MHz, CD₂Cl₂, 25 °C, δ): +134.8 (q, ²J_CF = 28.0 Hz, 2/5-C), +133.7 (q, ³J_CF = 3.7 Hz, 3/4/6-C), +132.3 (q, ²J_CF = 31.0 Hz, 2/5-C), +127.1 (q, ¹J_CF = 275 Hz, 2/5-CF₃), +126.3 (br.m, 3/4/6-C), +124.9 (q, ¹J_CF = 273 Hz, 2/5-CF₃), +121.6 (q, ³J_CF 3.9 Hz, 3/4/6-C); ¹⁹F NMR (470.67 MHz, CD₂Cl₂, 25 °C, δ): −59.3 (d, J_FH = 7.6 Hz, 9F, 2-CF₃), −63.6 (s, 9F, 5-CF₃). Elemental analysis (calc. for C₂₄H₁₀B₁F₁₈Na): C 42.93 (42.76), H 1.61 (1.50).

Acknowledgements

G. G. W. and A. E. A. thank the Royal Society for financial support by University Research Fellowships. E. J. L. thanks the EPSRC for financial support by a DTA studentship under grant number EP/J500409. The research leading to these results has received funding from the European Research Council under the ERC Grant Agreement no. 307061 (PiHOMER).

We thank Dr C. MacDonald (University of East Anglia) for assistance with the ¹H and ¹⁹F DOSY NMR experiments.

We acknowledge the use of the EPSRC funded National Chemical Database Service hosted by the Royal Society of Chemistry, and the EPSRC UK National Mass Spectrometry Facility (NMSF) at the University of Swansea. We thank the EPSRC UK National Crystallography Service at the University of Southampton for the collection of the crystallographic data.³⁰ We thank the Research Computing Service at the University of East Anglia for access to the high performance computing cluster.

References

1. G. C. Welch and D. W. Stephan, J. Am. Chem. Soc., 2007, 129, 1880–1881.
2. D. W. Stephan, Dalton Trans., 2009, 3129–3136.
3. D. W. Stephan and G. Erker, Angew. Chem., Int. Ed., 2010, 49, 46–76.
4. D. W. Stephan, Comp. Inorg. Chem. II, 2013, 1, 1069–1103.
5. D. W. Stephan and G. Erker, Chem. Sci., 2014, 5, 2625–2641.
6. D. W. Stephan, Acc. Chem. Res., 2015, 48, 306–316.
7. D. W. Stephan and G. Erker, Angew. Chem., Int. Ed., 2015, 54, 6400–6441.
8. P. A. Chase, G. C. Welch, T. Jurca and D. W. Stephan, Angew. Chem., Int. Ed., 2007, 46, 8050–8053.
9. K. Chernichenko, Á. Madarász, I. Pápai, M. Nieger, M. Leskelä and T. Repo, Nat. Chem., 2013, 5, 718–723.
10. H. Wang, R. Fröhlich, G. Kehra and G. Erker, Chem. Commun., 2008, 5966–5968.
11. T. Mahdi and D. W. Stephan, J. Am. Chem. Soc., 2012, 136, 15809–15812.
12. D. J. Scott, M. J. Fuchter and A. E. Ashley, J. Am. Chem. Soc., 2014, 136, 15813–15816.
13. G. Ménard, T. M. Gilbert, J. A. Hatnean, A. Kraft, I. Krossing and D. W. Stephan, Organometallics, 2013, 32, 4416–4422.
14. Z. Lu, Y. Wang, J. Liu, Y. Lin, Z. H. Li and H. Wang, Organometallics, 2013, 32, 6753–6758.
15. R. L. Melen, Chem. Commun., 2014, 50, 1161–1174.
16. S. C. Binding, H. Zaher, F. M. Chadwicka and D. O'Hare, Dalton Trans., 2012, 41, 9061–9066.
17. D. J. Scott, M. J. Fuchter and A. E. Ashley, Angew. Chem., Int. Ed., 2014, 53, 10218–10222.
18. T. J. Herrington, A. J. W. Thom, A. J. P. White and A. E. Ashley, Dalton Trans., 2012, 41, 9019–9022.
19. J. M. Farrell, J. A. Hatnean and D. W. Stephan, J. Am. Chem. Soc., 2012, 134, 15728–15731.
20. E. J. Lawrence, T. J. Herrington, A. E. Ashley and G. G. Wildgoose, Angew. Chem., Int. Ed., 2014, 53, 9922–9925.
21. G. Ménard and D. W. Stephan, Angew. Chem., Int. Ed., 2012, 53, 8272–8275.
22. E. R. Clark and M. J. Ingleson, Angew. Chem., Int. Ed., 2014, 53, 11306–11309.
23. A. M. Chapman, M. F. Haddow and D. F. Wass, J. Am. Chem. Soc., 2011, 133, 18463–18478.
24. X. Xu, G. Kehr, C. G. Daniliuc and G. Erker, J. Am. Chem. Soc., 2015, 137, 4550–4557.
25. E. J. Lawrence, V. S. Oganesyan, D. L. Hughes, A. E. Ashley and G. G. Wildgoose, J. Am. Chem. Soc., 2014, 136, 6031–6036.
26. E. J. Lawrence, R. J. Blagg, D. L. Hughes, A. E. Ashley and G. G. Wildgoose, Chem. – Eur. J., 2015, 21, 900–906.
27. E. L. Kolychev, T. Bannenberg, M. Freytag, C. G. Daniliuc, P. G. Jones and P. D. M. Tamm, Chem. – Eur. J., 2012, 18, 16938–16946.
28. S. M. Cornet, K. B. Dillon, C. D. Entwistle, M. A. Fox, A. E. Goeta, H. P. Goodwin, T. B. Marder and A. L. Thompson, Dalton Trans., 2003, 4395–4405.
29. W. V. Konze, B. L. Scott and G. J. Kubas, Chem. Commun., 1999, 1807–1808.
30. S. J. Coles and P. A. Gale, Chem. Sci., 2012, 3, 683–689.
31. E. J. Lawrence, V. S. Oganesyan, G. G. Wildgoose and A. E. Ashley, Dalton Trans., 2013, 42, 782–789.
32. D. H. McDaniel and H. C. Brown, J. Org. Chem., 1958, 23, 420–427.
33. T. A. Rokob, A. Hamza and I. Pápai, J. Am. Chem. Soc., 2009, 131, 10701–10710.
34. I. B. Sivaev and V. I. Bregadze, Coord. Chem. Rev., 2014, 270–271, 75–88.
35. Z. M. Heiden and A. P. Lathem, Organometallics, 2015, 34, 1818–1827.
36. G. R. Fulmer, A. J. M. Miller, N. H. Sherden, H. E. Gottlieb, A. Nudelman, B. M. Stoltz, J. E. Bercaw and K. I. Goldberg, Organometallics, 2010, 29, 2176–2179.
37. D. H. Wu, A. D. Chen and C. S. Johnson, J. Magn. Reson., Ser. A, 1995, 115, 260–264.
38. L. J. Bourhis, O. V. Dolomanov, R. J. Gildea, J. A. K. Howard and H. Puschmann, Acta Crystallogr., Sect. A: Fundam. Crystallogr., 2015, 71, 59–75; O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard and H. Puschmann, J. Appl. Crystallogr., 2009, 42, 339–341.
39. L. Palatinus and G. Chapuis, J. Appl. Crystallogr., 2007, 40, 786–790; L. Palatinus and A. van der Lee, J. Appl. Crystallogr., 2008, 41, 975–984; L. Palatinus, S. J. Prathapab and S. van Smaalen, J. Appl. Crystallogr., 2012, 45, 575–580.
40. G. M. Sheldrick, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 2015, 71, 3–8; G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystallogr., 2008, 64, 112–122.
41. S. Lancaster, ChemSpider Synthetic Pages, 2003,  10.1039/SP215, http://cssp.chemspider.com/215; T. E. Krattf, Boulder Scientific Company, Eur. Pat, 1186608, 2002; R. J. LeSuer, C. Buttolph and W. E. Geiger, Anal. Chem., 2004, 76, 6395–6401.
42. DigiElch-Professional, ElechSoft, Kleinromstedt, Germany.
43. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, J. E. Peralta Jr., F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, revision A.02, Gaussian, Inc., Wallingford, CT, 2009.
44. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
45. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter, 1988, 37, 785–789.
46. P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, 28, 213–222.

---

Footnote

† Electronic supplementary information (ESI) available: Comprising: annotated ORTEP plot of the X-ray crystal structure of 3; cyclic voltammograms of B(C₆F₅)₃ under corresponding conditions; further computational data. CCDC 1061234. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5dt01918d

---